FAANG Stock News Sentiment Analysis

Introduction

This project focuses on analyzing the sentiment of news articles related to FAANG (Facebook, Apple, Amazon, Netflix, Google) stocks. The goal is to understand the overall market sentiment surrounding these companies and its potential impact on stock prices. By leveraging sentiment analysis techniques, we aim to extract insights from textual data and identify trends and patterns in investor sentiment.

Analysis Components

• Data Preprocessing: Cleaning and preprocessing textual data to remove noise, such as stopwords, punctuation, and special characters. Tokenization and lemmatization techniques may also be applied to normalize the text.

• Sentiment Analysis: Utilizing sentiment analysis algorithms to classify news articles into positive, negative, or neutral sentiment categories. Techniques such as bag-of-words, TF-IDF, or pre-trained language models like BERT may be used for sentiment classification.

• Statistical Analysis: Conducting statistical tests to analyze the accuracy of the model.

  Technologies Used

• Python

• Pandas

• Scikit-learn

• NLTK

• Matplotlib

• Seaborn

Model

Random Forest Classifier: Random Forest is utilized as a classification model to classify sentiment categories based on the textual features extracted from news articles.

Repository Structure

• Dataset: 'FAANG_STOCK_NEWS_Dataset.csv' contains raw data files downloaded from Kaggle.
• Notebooks: 'FAANG_Stock_News_Analysis.ipynb' contains python notebook with code for data preprocessing , wrangling and sentiment analysis.
• README.md: This document providing an overview of the project, instructions, and guidelines for contributors.

Usage

1. Clone the repository to your local machine.
2. Install the required dependencies and libraries.
3. Explore the notebooks and scripts to understand the sentiment analysis process and run the code as needed.
4. Contribute to the project by adding new sentiment analysis techniques, improving existing models, or enhancing visualizations.